Sie sind auf Seite 1von 21

Classical Quarterly 50.2 563-583 (2000) Printed in Great Britain




Very soon after I began working on the identity of magic-workers in classical antiquity, I realized that it was necessary to come to terms with a thesis about depictions of erotic magic-working in Greek and Roman literature. It asserted that male writers engaged in a systematic misrepresentation of the realities of magic- working in portraying erotic magic as an exclusively female preserve; the reality was that men were the main participants in this form of magic-working. The thesis is based on the supposition that the truth about erotic magic and the people who performed it is to be found in the formularies or spell-books preserved in papyrus and in defixiones. These two sources of information are said to show us that erotic magic was performed by men and not by the women who are the persons depicted engaging in love-magic in literature. The scholar who first presented the thesis was the late John Winkler. A version of it is to be found in Fritz Graf's general account of Greek and Roman magic. 1 There is agreement over what are taken to be the facts, but views diverge over their interpretation. Winkler appeals to the Freudian notion of denial and transference to offer an explanation not only of the discrepancy between life and literature, but of what he took to be the belief held by the young men who cast erotic spells that the girls who were the objects of their spells were as sexually eager as they were: men, when overwhelmed by sexual desire for unattainable women, through a process of denial transfer that feeling to women, whether old or young, whom they

Abbreviations DTAud A. Audollent, Defixionum tabellae quotquot innotuerunt tarn in graecis Orientis quam in totius Occidentis partibus praeter atticas in Corpore inscriptionum atticarum editas (Paris, 1904) DTWu IG HI (3) = Appendix continens defixionum tabellas in attica regione repertas, ed. Richard Wunsch (Berlin, 1897) PGM K. Preisendanz, Papyri graecae magicae: die griechischen Zauberpapyri, 2nd edn rev. by A. Henrichs (Stuttgart, 1973) SupplMag Supplementum Magicum I, II, ed. Robert W. Daniel and Franco Maltomini, Abhandlungen der rheinisch-westfdlischen Akademie der Wissenschaften, Sonderreihe Papyrologica Coloniensia 16.1-2 (Opladen, 1990)

1 Winkler made his case in his The Constraints of Desire: The Anthropology of Sex and Gender in Ancient Greece (New York/London, 1990), 71-98, republished in an abbreviated form as 'The constraints of Eros', in C. A. Faraone and D. Obbink (edd.), Magika Hiera (New York, 1991), 214-43. Graf's book was originally published in French as La Magie dans I'antiquite greco- romaine (Paris, 1994), then translated into English as Magic in the Ancient World, tr. Franklin Philip (Cambridge, MA, 1998). There is, however, a revised edition in German: Gottemahe und Schadenzauber: die Magie in der griechisch-romischen Antike (Munich, 1996). The English translation will be cited for the sake of convenience. The thesis has encountered some opposition:

the authors, names not given, of the chapter on erotic binding-spells in Curse Tablets and Binding Spells from the Ancient World, ed. J. G. Gager (New York, 1992), 80; D. Montserrat, Sex and Society in Graeco-Roman Egypt (London, 1996), 187; and D. Ogden, 'Binding spells: curse tablets and voodoo dolls in the Greek and Roman worlds', in The Athlone History of Witchcraft and Magic in Europe II: Ancient Greece and Rome (London, 1999), 63-7 hold that Winkler has overstated his case.

Downloaded from https:/ Swansea University, on 22 May 2017 at 19:47:47, subject to the Cambridge Core terms of use, available at https:/



fondly imagine suffer the same intense sexual longings as themselves. 2 The long and the short of Winkler's thesis is that the male authors who portray female erotic magic-working are all in their own way just as much frustrated sexual fantasists as the unhappy young men with no outlet for their sexual longings who perform magical rituals directed at girls asleep in their beds onto whom they project their own tortured feelings. The women are unobtainable, because they are maidens of the kind who 'are constantly guarded and watched by their own families and by all the neighbours'. 3 Graf accepts Winkler's basic premiss of a discrepancy between literature and life; 4 his explanation is somewhat different, but similar in one highly significant respect:

both scholars believe the spells were directed by young men at young women, securely guarded within the bosoms of their families. 5 In Graf's view, the key to understanding what is going on is competition, that is, a jockeying for social position and advancement. The erotic binding-spells are then best understood as a manifestation of the same competitive impulse that gives rise to the binding-spells aimed at rivals in commerce or in the same calling. In the case of erotic binding-spells, their aim 'is almost always permanent union, that is, marriage', leading to social advancement. 6 In support of this last point, Graf cites the final phrase of one of the spells in PGM IV, which, on his interpretation of it, means that what is sought is 'a woman for one's whole life'. 7 Graf does concede that if marriage is the ultimate aim of erotic spells, then it is curious that no spells have been found aimed at the person who on his understanding of the social situation underlying the spells would in fact decide whom the woman married, the father. 8 Graf's explanation of why in literature it is 'always the women who perform erotic magic' is that such tales take magic out of the sphere of men, where it ought to have no place, since a real man does not use magic to attain his ends. 9 If I understand the explanation, what it means is that men make up stories about women practising erotic magic, precisely because they know that it is men who in the main engage in such conduct; they do so to show their disapproval of such a path of conduct and to impress on their readers the womanishness and weakness of such behaviour. That is tantamount to saying that there is a didactic and moral purpose to stories in which women engage in love-magic: they warn men to avoid it, since it belongs to the realm of women. The propositions on which the thesis rests need to be subjected to scrutiny. They are:

(i) the recipes in the magical formularies take it for granted that it will be men who will

2 Winkler (n. 1), 90. Endorsed by C. A. Faraone, TAPA 119 (1989), 149-61; S. Johnston, TAPA 125 (1995), 179, n. 3; and D. Martinez in Ritual Power in the Ancient World, ed. M. Meyer and P. Mirecki (Leiden, 1995), 354-5; with qualifications by H. Versnel in Ansichten griechischer Rituale: Geburtstag-Symposium fur Walter Burkert, ed. F. Graf (Leipzig, 1998), 257-8 and C. A. Faraone, Ancient Greek Love Magic (Cambridge, MA, 1999), 82-5.

3 Winkler (n. 1), p. 90.

4 Graf (n. 1), 185: 'In Theocritus as well as Vergil, or in the elegiac poets, and generally in the great majority of the literary texts, it is women who practise magic, whether erotic or of another kind. This situation amounts to an astonishing .reversal of what we find in the epigraphic texts and the recipes on papyrus.'

5 Ibid., 186.

6 Ibid., 186. Ogden (n. 1), 66 seems to favour Graf's position, but is ultimately non-committal on the issue.

7 404—5: Kai TO. a<j>po8iotaKa iavrrjs Xpovov TOV ala>vos.

e/creAtaij i) Sztva fier'


Seiva, els TOV airavra

8 Graf

(n. 1), 188.

» Ibid., 199.

Downloaded from https:/ Swansea University, on 22 May 2017 at 19:47:47, subject to the Cambridge Core terms of use, available at https:/



perform erotic magic; 10 (ii) erotic deflxiones were cast by men and aimed at women; (iii) the women who are the targets of love-magic are carefully guarded maidens; (iv) the motives of the men who employ love-magic are either to enjoy the sexual favours of an otherwise unattainable woman or to better themselves by securing a good marriage; (v) in literature it is exclusively or almost exclusively women who are portrayed engaging in love-magic. In what follows erotic magic should be taken to mean any form of magic intended to manipulate the sexual behaviour of others. Erotic magic on this understanding encompasses not only magical rites aimed at bringing a woman or man to the bed of the person performing them, but also spells cast by persons motivated neither by sexual longing nor jealousy. A spell cast by a prostitute and intended to turn clients away from a rival is an erotic spell, even though the motive for casting the spell is not sexual.


The surviving magical formularies containing recipes for erotic spells are with one exception the products of Late Roman Egypt, although some of the spells in them may be a good deal older." Whether all of the erotic spells in the formularies had their origin in Egypt and were devised by magicians in that land is not clear. What is

certain is that such spells circulated throughout the Mediterranean. 12 In analysing erotic spells in formularies it is necessary to bear in mind that the spells were devised by professional magicians and were then copied out from formularies not by the persons who speak in the first person in them, but by scribes-cum-magicians

commissioned to perform that task.

the spells are written is, accordingly, a not very satisfactory guide to the state of mind of young Egyptian, let alone Mediterranean, males. The scribes-cum-magicians who used the formularies nonetheless adapted the recipes in them to fit the circumstances of their clients, if not their psychological needs. What that means amongst other things is that a spell in a formulary in which a man is the agent or dejigens and a woman the passive party or defixa could be adjusted mutatis mutandis to suit the needs of a woman seeking a man or a man pursuing another man. PGM XXXIIa and LXVIII are a case in point. They are both defixiones written on papyrus that come from Hawara and belong to the second or third century A.D. The same hand wrote both and used the same formulary, but made appropriate adjustments: PGM XXXIIa is a homosexual spell directed by one man at another; PGM LXVIII was written for a woman whose aim it was to draw a man to her. 14 The recipe in the formulary that the scribe used must almost certainly have taken the form:


The urgent and very often violent tone in which



TW}><JJV avrihiKOS





KO.1 <fiv)(rjv Tijy

Seiva rjv CTCKCV 17 Stlva KTX. It gave rise, on the one hand, to: ws 6 Tv<j>wv


, ovjws

tfrvxyv aiiTOV Avcovelov, ov ereKev ' EXevrj KTX. (PGM XXXIIa. 1-4) and,

10 Winkler (n. 1), 90 does concede that there are two exceptions: PGM 1.98, IV.2089.

" The exception is SupplMag 72, a formulary from the Augustan period.

12 On this phenomenon, which he describes as a 'community of superstition in the oikoumene in the time of the Empire', see D. R. Jordan, Hesperia 63 (1994), 123-5.

13 The hands in which the spells are written show that professionals are at work. Whether the professionals were lector-priests is another matter. R. Ritner, 'Egyptian magical practice under the Roman Empire: the demotic spells and their religious context', ANRW2.\%.S (1995), 3354-8 and D. Frankfurter, Religion in Roman Egypt (Princeton, 1998), 224-37 assume that they were.

14 I am much indebted to David Jordan for drawing my attention to the pair.

Downloaded from https:/ Swansea University, on 22 May 2017 at 19:47:47, subject to the Cambridge Core terms of use, available at https:/



the Other, to : ws










ereKev Zoia\i\x.t), ini avrrj[v 'Ep]t[eav KTX. (.PGM LXVIII. 1-3). It is worth observing

in general that the spells cast in Late Roman Egypt by women to lure men or other

women to them do not greatly differ in form from binding-spells directed by men at women. Almost all the expressions found in them have their counterparts in spells aimed by men at women.

That the recipes can be and were adjusted suggests that not too much weight is to be given to the fact that in most model-spells a man is the agent and a woman the object. The predominance of the pattern tells us very little about the assumptions made by those who composed the spell. In all probability, the existence of the pattern has its origins in the precedence accorded the masculine grammatical gender in Greek. That supposition is supported by a formulary (PGM XXXVI, fourth century A.D.) in which

a spell, specifically advertised as being good no matter whether it is a man acting against a woman or a woman against a man, is laid out as if the woman was the victim and the man the one who inflicted the suffering. We are first told that it is a most

excellent fire-spell than which none is mightier (efxirvpov fieXnoTov, ov ixit,ov ovSev 70) and then that it brings men to women and women to men and causes maidens to

avSpeaiv Kal

come bounding out of their homes (ayi Se dvSpas ywe^lv Kal yvveKas

napOevovs eKn-qSav oiKoOev 71). But in spelling out what the operant should do the text proceeds as if it were only a question of a man attracting a woman:

Xajitbv ^aprijv

Kadapov ypd<f>e aip.ari

fiaXwv ovaiav i^s dtXus yvvaiKos.

ovita TO. imoKip.eva ovofiara

Kal TO £WI8IOV, Kal


Take a clean sheet of paper and write the following names and inscribe the following figure using the blood of an ass and throw on to it something belonging to the woman you want.

The performer is then told to put gum on the paper, so that it may be attached to the sudatorium in the baths. He is advised that he will be amazed, but warned that he should watch out, lest he receive a heavy blow. The text of what is to be written on the

sheet of paper follows. It is one of the standard prayers found in such spells. The prayer

is phrased as if it too concerned only a man drawing a woman to him:

OVTWS Kal 17 '{"'XV' V KapSia rij j Seiva, ijy er€K(v r) heiva,

€coy av eXOrj <f>i\ovoa ip.€ TOV Selva Kal Tqv dTqXvKTjv (j>vaiv TT\ dpaeviKJj fiov KoXXr/arj.

d>S vp.(is KaUade Kal nvpovade,


As you burn and are consumed with fire, let the same thing happen to the heart and soul of such-and-such a woman, whom so-and-so bore, until she comes in love to me being so-and-so that she may join her female parts to my male parts.

There is a similar pattern observable in a recipe for an attraction-spell in a formulary (PGM IV1495-1595). It requires the burning of myrrh, which is to be invoked as if a demonic force that roasts and compels persons (generalizing masc. pi.) who deny Eros

t o fee l lov e (17 <f>pvyovaa Kal avayKa^ovaa

</>iXeiv rovs




"Epuna. 1500-2). Despite the generalizing use of the masculine gender, the model for the spell itself is written as if for a man who would dispatch the burning myrrh to enter and scorch a woman, until she was brought to his side (1505-95). PGM XXXVI is by no means the only formulary to contain a recipe that explicitly says it can be used to draw a woman to a man or vice versa. Another does the same thing, recommending that the name of the Sun be uttered three times, and boasts that

it brings a woman to a man and a man to a woman in a way that is simply amazing

(ayei yvvaiKa dv8pl Kal avSpa yvvaiKi ware 9avp,doai PGM XIII.24—6). Another spell-book has a recipe entitled IT[d]pe8pos "Epws, in which Eros is summoned to be

Downloaded from https:/ Swansea University, on 22 May 2017 at 19:47:47, subject to the Cambridge Core terms of use, available at https:/


the helper or familiar of the performer (PGM XII. 14-95). In the second and third prayers prescribed, to be uttered over a complex sacrifice, the performer is to call on the Kosmokrator to turn all women to him in love, giving his name as so-and-so, if he is a man, or so-and-so, if a woman (TTOI-^OOV oTpe'<f>eo0[ai Trav^Tas avQpwirovs TE Kal TT&oas ywaiKas inl \e\pwTa fiov, TOV Seiva (rj rrjs Seiva) 60-1). One of the recipes in PGM IV has a different pattern from the one just noted: before the model-incantation and any accompanying actions needed to make it effective are set out, the application of the spell is described; in three instances the masculine singular gender is used to refer to the persons it is said to draw, but in spelling out what has to be said to make it effective the example given is always of a woman with name to be supplied being drawn to a man with name to be supplied. 15 There is yet another spell in the same formulary (PGM IV) which, although it does not specify the sex of the party whom it draws, renders bed-ridden through sickness, sends a dream to, binds down or summons a dream from (dyei 8e Kal KaraKXLvei Kal oveipoTrofinei xal Karexei xal oveipanrjTei 2076-8), does, however, command the demon summoned as helper to go to where the woman or the man who is the object of the magic dwells, but proceeds only to bid the demon to bring a woman, name to be supplied, to a man, name to be supplied (nopevov, OTTOV KdToi/cef 178c (77 oaSe), Kal a£ov ainrjv TTpos ifxe TOV Seiva 2089-91). Finally, there is a recipe in a very early formulary of the Augustan Age that would most appropriately be used by a woman, but which allows for the possibility of its use by a man; grammatical gender is hopelessly confused throughout. The spell consists in anointing the face with myrrh, while invoking myrrh as the myrrh with which Isis anointed herself when she went to the bosom of Orisis and which gave her x<*P's- on that day (SupplMag 2.72.ii.l—25). Since the mythical paradigm has a woman anoint- ing herself to make herself more attractive in the eyes of a man, the presumption must be that the spell was intended primarily for the use of women. Yet, in the instructions for the spell, participles in the masculine nominative are used (apas, Xeycov 1; XafSwv 4), but a second invocation in the spell of Isis covers the possibility that either a man or a woman might be its object (TOV Seiva rj rr/v Seiva 7). On the other hand, in the rest of the prayer that is to be uttered it is a man who is being pursued (nepl TOV Seiva 11; avros Se /j,e <f>evyei. 12; TOV Seiva 21, 22). The simplest explanation for the confusion is that recipes for spells tend to be written in the masculine gender, even when they would most appropriately be used by a woman. The main conclusion to be drawn from our examination of the formularies is that the authors of such spells do not necessarily assume that only women are to be the object of their spells. The likely explanation is that the use of the masculine gram- matical gender is merely a convention for writing out as economically as possible spells that may be used equally well by either sex. It is possible to conclude with some confidence that in Late Roman Egypt those who copied out the model-spells for spell-books did so in the belief that some of their spells might be used by women. They

15 1: (a) Sl<f>os AapSdvov.


Kal ayei ijivxqv dvriKpvs

0$ av BeXrjs, Xeywv TOV

\6yov Kal on KXIVW rf/v I/IVXV TOV Seiva (1716-22); (b) eTriarpeifiov TIJV <l>vxqv TTJS Seiva els ip.e TOV Seiva, "va fie 4"-Xfj, "va fiov epa, iva fioi Soi TO. ev rais xepolv eavrqs (1806-10) ; (c ) Kal eXdwv oifie eiy TT/V OIKIOV ij? jSouAei (1852—4). 2: (a) eoTiv yap Kaprepov Xiav xai avvTTe'pf}Xr)TOV, TTOIOVV irpos iravTas avdrjixepov (\%Ti—5);(b)efra avoi^as rqv Bvpav eupijcreis irapa. TCUJ Bvpais TJV Xeyeis (1906-8); (c) a/jov ftoi TTJV Seiva T17? Seiva (1915). 3: (a) OKexrr) em8vfji.aTos aeXr/viaKov ayovaa aoxerovs Kal avovaidoTOVS p.ovorniepovs (2441-4); (b) PdSiaov irpos TTJV Seiva Kal jSdaTofov avrrjs TOV VTTVOV Kal 80s avTrj Kavaiv >]>vxf)s (2446-8).

Downloaded from https:/ Swansea University, on 22 May 2017 at 19:47:47, subject to the Cambridge Core terms of use, available at https:/




were, as it turns out, not deluded: women in Late Roman Egypt, as we shall see, did engage in erotic magic.


The next question to be addressed is whether the users of the spells were lovelorn young men unable to gain access to maidens whose chastity was carefully protected by their watchful families and neighbours or were perhaps young men, not so lovelorn, but looking to better themselves by an advantageous marriage with girls from good families. If we look at the recipe in PGM IV entitled i>iXTpoKard8eafios and the five defixiones which used as their template either the recipe in PGM IV or one very like it, the indications are that the men for whom the recipe was copied out were intent neither on seducing sheltered virgins nor on securing a wife of good family, but had rather different goals in mind. All five copies depart from their model or (more likely) models in ways that are quite revealing. Two of the defixiones come from Oxyrhynchos and are directed at the same woman {SupplMag 49, 50), one is from a cemetery at Arsinoe (SupplMag 46), one perhaps from Antinoopolis (SupplMag 47), and the last is of unknown provenance (SupplMag 48). The model- spell calls on the demon-helper to betake himself to every place, every street or block and every house, there to bring and bind down the woman (v-naye els -n&vra TOTTOV

Kal els


KO.1 els


oiVi'av PGM

IV.348-9). All of the


except the pair from Oxyrhynchos follow the model more or less exactly. The two spells from Oxyrhynchos, which are aimed by a certain Theodoras at a woman named Matrona, broaden the scope of the search even further by adding to the list 'to every tavern' (els TTSLV KanrjXiov SupplMag 49.19, 50.19). Substantial variation occurs a little further on in the spell. The model-spell prays that the woman not enjoy vaginal or anal intercourse and that she should do nothing else contributing to pleasure with


Other ma n



fi-q nvyioBr/Tco,

fjLTjSe Trpos rjhovrjv


dXXov dvSpos PGMTV.351-2). One of the Oxyrhynchos-spells lacks this tricolon; the other adds fellation (/HI) [Xai]Kdarf) to the list of sexual activities the woman is not to perform and expands and clarifies the final element: the woman is not to fulfil the pleasures of love with another, nor have union with any other man (fi[r) i]K -qSovrjs






fj.r) [dA]Aaj





detail from the model (SupplMag 47.8-9); the one from Arsinoe adds fellation to the sexual activities that are specified (jiri XeiKaor) SupplMag 46.9). n The spell of unknown provenance adheres to the model, adding only that the woman should do nothing that contributes to the pleasure of another youth or another man (erepw veavtoKw rj aXXw dvSpi SupplMag 48.8). Whoever it is who introduces the departures from the model does so because he or she wishes to be precise where the model is too general or because he or she has a particular concern in mind. One imagines that the scribes-cum-magicians who copied out defixiones did not make substantial alterations on their own authority, but only after scribe and client had

The spell from the area of Antinoopolis does not depart in any significant

16 David Jordan per litteras tells me that, so far he can see, the form of XaiKa^etv visible in line 22 is the passive Xai]Kaadij. If that is the case, I would assume that the pattern of passives established byfiivrjOijand nvyiaOfi has led to the form.

Cf. SupplMag

38.3—5: iva

p.rj SvvTjd^s

€T€pq) av&pl


TTWTTOTZ ju-^re f$iv7]9ijvaL

fj.rJTC TTvyiaBijvai



/xrjSe KO.6' rjhovrjv woiijcnjs JJXB' eralpw


Downloaded from https:/ Swansea University, on 22 May 2017 at 19:47:47, subject to the Cambridge Core terms of use, available at https:/


conferred. It is possible, and indeed likely, that in some cases the person who writes out the binding-spell is also the person who casts it, but the style of writing found in Egyptian defixiones points very strongly to their having been copied out by scribes well-versed in inscribing such texts on lead. The variations on the model-text in PGM IV tell us how people interpreted it and for what kind of situations they understood it to be intended. But whatever the differences, the four different persons who used it were not trying to secure the undivided attentions and affection of a maiden of good family, but were more concerned to bring to their beds for their own exclusive enjoyment women who had been known to gratify the sexual needs of other men. It is not readily apparent why a man lusting after a maiden, even allowing for the rather greater degree of freedom allowed Egyptian women, would feel it necessary to avert the danger of her satisfying the varied sexual tastes of other men and in two cases felt that it was relevant to add fellation to the list. Nor again is it obvious why a demon should have been expected to find the woman he is sent in search of in a tavern, which is to say, in a centre for sexual commerce, if she was a carefully nurtured virgin living in the bosom of her family, or for all that, why someone should have thought it prudent to make sure that the girl he longed for was not servicing the sexual appetites of youths and men. There is, in any case, an explicit clause in the model-spell that tells against the possibility that the ultimate purpose of the spell was to secure a marriage with a virgin: the demon is asked to keep the woman from food and drink and not allow her to have a pleasurable experience with any other man, not even with 'her own man' (/*.•>) idajjs TTJV Seiva aXXov dvSpos Treipav Xafieiv irpos •qSovr/v, JJ,T)8£ ISIOV dvSpos PGM IV.374-5). 18 Who is the iSios dv-qp with whom the woman is to have no experience? It could be her husband, but it is equally likely that the phrase covers a less formal relationship: it may mean only the man with whom the woman is currently living. Whoever it is who is intended, it is fairly clear that the model-spell does not have in mind a maiden living at home under the protection of her family. It is tempting at this point to conclude that those who used the model-spell tended to be men intent on making their own exclusive sexual property prostitutes or women who were prepared to bestow their sexual favours on others. Nonetheless, proponents of the thesis that marriage was the goal that users of such spells had in mind will point to the wish in the model-spell, followed with variation in its adaptations, that the man might be united to the woman for the rest of his or her life. In the spell from Arsinoe (SupplMag 46), for instance, the man Posidonius wishes that the woman Heronous may not be parted from him until death (K-CU dSiaxwpt-oTov fiov avTrjv Troiija^ Bavarov 25-6) and wishes that he may have her for all the time of his life (iV

avrr/v 'HpOvovv

In the model-spell, on the other hand, the wish is that the man's head, lips, and belly may form a bond with the woman's, that his thighs may come close to hers, that his black (that is, pubic hair) may be joined to her black and that she may perform the deeds of love with him for all the time of life (/cat KefaXrjv K€<f>aXfj KoXX-qor) KCLL

XEi'Aea ^eiAeat avvaxf>r) Kai yaaripa yaarpl KoXXr/arj KOLI firjpov y.t]p<^> neXdaji KCLI


Is TOV a-Travra XP° VOV TTJS ^oifjs fiov 26-8).

TO jU.eAavi avvdpixoarj



d<f>po8toio.Ka eKTeXear) r/ Seiva





els TOV dnavTa XP° VOV T0 " aldivos PGM IV.400-6). The rather more chastely expressed wishes of Posidonius might be construed to come to very much the same

18 That clause is reproduced in a simplified form in only two of the adaptations, but in them no mention is made of any experience being had with the person called the ZSios avr/p (SupplMag 46.21,47.21). He presumably did not exist and was left out.

Downloaded from https:/ Swansea University, on 22 May 2017 at 19:47:47, subject to the Cambridge Core terms of use, available at https:/



thing as the clause 'until death do us part' in the Christian marriage-ceremony, which would, if true, point to its being marriage that Posidonius had in mind. It could be argued, in any case, that there is nothing incompatible in longing for unceasing sexual union and for marriage that ends only in death. But to interpret either version of the spell as expressing a wish for marriage would be a mistake. What the spell for Posidonius is intended to procure is that Heronous shall have no life of her own apart from Posidonius and that she shall be utterly subordinate to him sexually for the rest of her or his days. The crucial word here is tnroT€Tayfievrjv (27). It also occurs in the same context in the binding-spell from Antinoopolis or vicinity, but with the added qualification that the woman shall love and crave the man and tell him all that she has in her mind (vnorerayfievrjv els rov




£,a)?js fJ,ov <f>i\ovadv fj,€, epwa\a\v

fj-ov, Xeyovaav


a ex^i


vu> SupplMag 47.26-7). 19 There is nothing touching about this last wish. The man is looking for complete domination over the woman: he does not want her to think of any other men. Any doubts about the connotations of the verb vnordaaeadai in erotic defixiones should be put to rest by a defixio from Hadrumetum in Tunisia (third century A.D.), in which a woman wants to make a man her partner in life and perhaps in marriage and to subordinate him to her, loving her like a slave and desiring no

other woman or maiden than herself (Troir/aov avrov

VTTOTeTaxOrjvai, p.-qhefxlav dXXt)\y\ yvvaiKa

d>s SovXov avryi


fxr)re napdevov

e-7n6vfj.ovvra DTAud

271.44-5). 20 As for the formula els rov diravra avrrjs xpovov, when used in defixiones, it does not reflect love eternal, but a desire to make the spell effective for the life of the interested parties 21

Some of the spells in the formularies are for debauching married women, or women in a permanent relationship with a man; others for making a woman forget her husband or current companion, her parents, children, and brothers and sisters (PGM IV2755-61, LXI.29-30). The assumption should not, however, be made that there was a hard and fast dividing line between women who were still closely involved with their families and women who were having relations with other men. A recipe at PGM IV274O-58 commands that if the woman lies holding another man to her bosom, she should thrust him away from her, having put all thought of children and parents from her mind (PGM IV2756-8). 22 The recipes in the formularies, therefore, do not on the whole seem to have been written for men who had marriage in mind. Even the term ydpos in such recipes is likely to refer not to marriage but to sexual union. 23 In the formulary PGM V, for example, there is a recipe with the phrase eav 8e yvvaiKa- 'omos fi-q yaftrjarj rov Seiva -q heiva.' (330-1), whose context points to sexual union rather than marriage. It would be extremely imprudent to deduce, from the stated goal of one erotic spell, the unstated purpose of all others, but there is a revealing recipe at PGM XXVI. 134-60, to

" Cf. PGM VII.610: Si6 a£arf p.oi avrfjv <f>\eyop.fin)v, vnoraaaofievrjv.

20 On the spells known as imoraKTiKa, whose express purpose was to make others totally and unquestioningly obedient, see T. Hopfner, Archiv Orientdlni 10 (1938), 135-46.

21 So P. Moraux, Une Deftxionjudicaire au Musee dIstanbul, Memoires de I'Academie royale de Belgique, Classe des lettres et des sciences morales et politiques, 2nd ser. 54.2 (Brussels, 1960), 55-6. Cf. PGM Vll.650-1, XVII.15-16.

22 Cf. the papyrus erotic spell SupplMag 45 (Assiut in Upper Egypt fifth century AD. ) with the same wishes (lines 46-50), but not, however, the mention of children.

23 Cf. L.Robert , Collection Froehner I: Inscriptions grecques (Paris, 1936), 18, n. I; id., RPhilAX (1967), 77-81; Emmanuel Voutiras, AI0NY20&QNT0Z FAMOI: Marital Life and Magic in Fourth Century Pella (Amsterdam, 1998), 55-6.

Downloaded from https:/ Swansea University, on 22 May 2017 at 19:47:47, subject to the Cambridge Core terms of use, available at https:/



make a woman suffer the usual torments, so that she may consent to love the man as would a woman available for sex (iva poi iirivevori em eVcnpamK-jj <f>iXia PGM XXXVI. 144-53). 24 My suspicion is that we can see here the market at which the erotic spells in the formularies were mainly directed: young men who wanted to monopolize the sexual activities of the kind of woman with whom they would not have considered entering into a formal contractual relationship. To sum up, while it may be very difficult to say who the ideal user of the erotic recipes in the formularies was, if such a creature existed, there is very little to support the belief that it was either a lovelorn young man yearning for an unattainable maiden or a calculating young man trying to secure a marriage with a girl of good family. Nor do the formularies give the impression that their primary market was young men attempting to seduce respectable women. No doubt a good deal of that went on, but there is little evidence of it in the formularies. The formularies seem to look to sexually active women who are either available or whom there is some hope of detaching from their present attachments. Who are these women and in what circles did they move in Late Roman Egypt? Conditions and customs in Egypt at that time were very different from those that prevailed in the small segment of Athenian society in the fifth and fourth centuries B.C. about which anything is known. Yet Winkler writes about the women who are the objects of Egyptian erotic spells as if they were the carefully protected daughters of well-to-do fourth-century Athenians or girls living in a village in Crete in the 1970s. He has more or less unquestioningly assumed that Mediterranean social life is an undifferentiated whole, unaffected by place or time. 25 His picture of a homogeneous whole, existing unchanged through time and space, seems to rest on a few studies in English, principally of modern-day Greek village-communities, very often isolated in the mountains. 26 It is dangerous to generalize about relations between men and women, but it is hard to avoid the impression that in the larger cities of Egypt women were allowed a good deal of freedom and that many women who were certainly not prostitutes enjoyed sexual relations with men outside of marriage, an institution which in Egypt was in any case somewhat more loose-knit than it was in classical Athens. 27 That impression is based on letters and legal documents. Sophronius in his hagiography of a harlot-saint tells of the saint as a twelve-year old girl fleeing the home of her parents to come to Alexandria, where she had lost her virginity and had for the next seventeen years continued to let men enjoy her body. It was not that she needed to earn money with her body—she often after all refused payment for the services she provided—lust and a desire for pleasure were what motivated her {V. Mar. Aeg. 18; PG 87(3).3710-12). 28 A

24 On the extension in meaning that iraipa undergoes in the Roman period, where it becomes a general word for a prostitute, see S. Leontsini, Die Prostitution imfriihen Byzanz (Vienna, 1989), 25-6; Montserrat (n. 1), 107-8.

25 For criticism of Winkler's monolithic view of the Mediterranean, see C. Paglia, Sex, Art, and American Culture (New York, 1992), 193-207.

26 The studies cited are: J. du Boulay, Portrait of a Greek Mountain Village (Oxford, 1974); id., 'Lies, mockery and family integrity', in J. Peristiany (ed.), Mediterranean Family Structures (Cambridge, 1976), 389-406; M. Herzfeld, The Poetics of Manhood: Contest and Identity in a Cretan Mountain Village (Princeton, 1985); id., Anthropology 9 (1985), 25-44.

27 On sex, marriage, and divorce in Late Roman Egypt, see R. Bagnall, Egypt in Late Antiquity (Princeton, 1993), 188-99; on marriage and divorce in Graeco-Roman Egypt, see Montserrat (n. 1), 80-105.

28 On the genre to which the Life of Mary of Egypt belongs, see B. Ward SLG, Harlots of the Desert (Oxford, 1987).

Downloaded from https:/ Swansea University, on 22 May 2017 at 19:47:47, subject to the Cambridge Core terms of use, available at https:/




good deal of exaggeration on Sophronius' part has to be allowed for, but that should not blind us to the very real possibility that he is describing a recognized pattern of behaviour and that there were twelve-year-old girls who ran away from home to Alexandria for adventure and who slept freely with men, not always taking payment for their services. There were no doubt other types of sexually active women who were neither prostitutes nor wives. Women who were divorced or separated from their husbands will have been one; widows will have been another. There were at least in the larger towns of Late Roman Egypt opportunities, both public and private, for men and women to initiate sexual relations. Clement of Alexandria assumes that grown women will drink in the company of men other than those in their own immediate family and that when drinking from the small-mouthed alabaster-vessel they favoured they will throw back their heads and excite the men present by deliberately exposing their necks and as much of the rest of their bodies as they can (Paed. 2.2.20-1,33, 1.168,176 Stahlin). Clement almost certainly describes fashions and customs prevalent in Alexandria in his own day, which is to say in the early third century A.D. The practices that he criticizes on the part of women are those of the well-to-do. The freedoms allowed rich women were not necessarily those granted other women in Alexandria, let alone women elsewhere in Egypt. Nor would it be wise to argue that if rich women were granted freedoms, a fortiori poorer women were granted still more. There are also indications in the spell-books that women, perhaps not of the exalted status of the women Clement criticizes, did drink with men. The recipes for the spell called a -rrorqpiov (PGM VII.385-9, 619-27, 643-51) or a TTOTiafj-a (PGM VTI.969-71) or in one case a <j>i\rpov norifiov (PGM XIII.319-20) presuppose men and women drinking together, since the spell is meant to be uttered by the man over the cup that he is then to hand to the woman whose love he seeks. Other spells were supposed to be intoned while kissing a woman, although the formularies do not say exactly under what circumstances they were to be performed (PGM VII.405-6, 661-2); that it was in a situation of some intimacy goes without saying. It may be that the women whom the authors of the drink-spells and the kissing-spells had in mind were not exactly the same sort of women whose presence at symposia gave Clement such cause for concern. What is clear at any rate is that in Egypt there were women who were not segregated from men and that the magic employed by men against them or by them against men was by no means always performed at a distance from the intended victim. Both parties might know each other quite well and already be enjoying sexual relations.

There were also opportunities for the sexes to meet in public places. There was, for instance, the practice widespread throughout the Roman Empire of mixed bathing in

the public baths.

the public baths and picking them up there (Paed. 3.5,1.254-5 Stahlin). Because of the opportunities for physical contact that the baths afforded, they were also suitable locations for certain types of magic-working. Epiphanius, writing towards the end of the fourth century A.D., recounts an incident in the baths at Gadara set in the reign of Constantine. The protagonist in it is the youthful Jewish patriarch of Tiberias, who had gone to Gadara from Tiberias with some companions to attend a festival. He had


Clement complains about well-to-do women bathing with men in

29 Mixed bathing in Rome: Plin. //.AC 33.153; Quint. 5.9.14; SHA Hadr. 18.11 (Dio Cass. 69.8.2); SHA Alex. Sev. 24.2; naked mixed bathing in Rome: Mart. 3.51, 72 (cf. 11.47); mixed bathing in unspecified locations: A. P. 9.621, 622, 783; mixed bathing naked, presumably in Alexandria: Clem Alex. Paed. 3.5.32, 1.254-5 Stahlin. See further, Garrett G. Fagan, Bathing in Public in the Roman World (Ann Arbor, 1999), 26-9.

Downloaded from https:/ Swansea University, on 22 May 2017 at 19:47:47, subject to the Cambridge Core terms of use, available at https:/



noticed in the baths an attractive young Christian woman. That had led to his delib-

erately rubbing his flank against hers in one of the chambers of the bath, probably the sudatorium. The woman had crossed herself, and God, to show his miraculous power, in spite of the sin the woman had committed in bathing in the company of men, had

caused the youth to fail in his attempt (Haer.

Epiphanius makes later about the protection the Sign of the Cross had afforded the young woman against the magic-working of the youth and his companions show that the youth was practising magic (Haer. Religious festivals would certainly have been one of the occasions on which young men and women came into contact with each other. Some idea of what went on on these occasions may be gained from a recently discovered sermon of Augustine in which he looks back to the Carthage of his youth: men and women mixed quite freely in all-night vigils; the young men took the opportunity afforded to make assaults on the chastity of young women; they would contrive to rub themselves against the women in the narrow and unsegregated passageway into the church; Augustine had himself been party to such doings; he remembers the improper songs to be heard at the feast of the Martyr Cyprian and the general wantonness of the proceedings. 31 In the Confessions, he is more specific and admits to having made successfully within the walls of a church arrangements for a sexual assignation (3.3.5). The girls who came to such vigils were not presumably the heiresses whom ambitious men sought to marry, but girls of relatively humble status who were not necessarily the social equals of the men who picked them up. These were the girls whom a man might make his concubine and then send away as Augustine did when an advantageous marriage became possible. 32 30 Some comments that


There are fifteen published erotic defixiones from Egypt directed by men at women. 33 One man directs a defixio at another (PGM XXXIIa). Two women, on the other hand, had erotic defixiones written for them to be directed at other women (PGM XXXII, SupplMag 42). Finally, there are two defixiones directed by women at men (PGM XV, XVI). There is no indication in any of the defixiones cast by men against women that marriage is what the men had in mind. So far as can be judged, most of the women at whom the spells are aimed are not closely guarded maidens, but are sexually available, although the concern that PGM XVIIa displays with overcoming a women's sense of shame suggests a woman with some pretensions to respectability. There are undoubtedly more defixiones cast by men than by women, but what this tells us is unclear, since the sample is not very large. For all we know, women may have had a preference for forms of erotic spell that leave no trace in the material record, or they may have lacked the confidence to approach the male scribes-cum-magicians who supplied defixiones, or they may have lacked the financial means to pay for the services of the scribes-cum-magicians. However that may be, what does emerge is that women did engage in erotic magic and that those who composed spells for them used the same formularies as were used for men.

30 For a description with illustrations of the baths at Hammat Gader (Gadara), see F. Yigiil, Baths and Bathing in Classical Antiquity (Cambridge, MA, 1992), 121-4.

31 F. Dolbeau, Augustin d'Hippone, Vingt-six sermons aupeuple d'Afrique (Paris, 1996), no. 5.5.

32 On concubinage, see P. Brown, Augustine of Hippo (London, 1967), 62.

33 PGM XVIIa-c, XlXa, XXXIX, LXVIII; SupplMag 38, 39,40,41,43,44,45,46, 47,48,49,


Downloaded from https:/ Swansea University, on 22 May 2017 at 19:47:47, subject to the Cambridge Core terms of use, available at https:/





From a cemetery or cemeteries outside Hadrumetum in Tunisia there are eight or possibly nine erotic defixiones?* In none of the six or seven spells directed at women does marriage seem to be the object. The text of one of the two defixiones com- missioned by women is quite informative {DTAud 271). The goal of the woman who has commissioned the spell, more explicitly and more fully expressed as the spell progresses, is to have the man filled with passion come to take her off to his house to live with him, and to have him subordinated to her as her slave. The man was obviously, in the eyes of the woman, something of a catch and sounds as though he represented economic security for her, if she could get hold of him. She wants him to be filled with passion and lust for her, not because she wants to have intercourse with him, but because that will keep his attentions from wandering and keep him loyal to her. In this she is quite unlike the men who have defixiones composed for them: their wish is to have women come to them in a state of fervid desire, so that copulation may take place forthwith. Her main goal is likely to have been economic security. Women are again greatly outnumbered by men in the binding-spells from Hadru- metum, but they do nonetheless cast such spells. Where the motive of the woman is discernible, it is to have the man keep her in his house in a permanent relationship. There is little evidence of such a desire on the part of the men who cast the spells.


There are three erotic binding-spells from Carthage. One of them asks that Successa be made to burn with love and desire for Successus {DTAud 227); from the names of the parties it sounds as if the pair were slaves. A more complex spell calls on a demon mighty amongst the Egyptians to bring a woman to a man to have intercourse with him (DTAud230A,B). The demon is, in addition, asked to drive the woman from her parents, her bedchamber, and from whomsoever she holds dear (a suis parentibus, a suo cubile et aerie quicumque caws habes 10-11). There is not enough information here with which to do much. The most interesting of the three is one that seeks to have a woman respond to the love and desire that a man called Martialis feels for her with a matching love and desire (DTAud 231). This is an extraordinary document, because the man acknowledges the feelings he has for the woman he is trying to attract. There are no parallels.


Because of the amount of archaeological activity that has taken place in recent times within its borders, Macedonia now boasts an increasingly large dossier of defixiones} 5 Most of them are early and come from the fourth or third centuries B.C. Two of them fall into the category of erotic magic. One of them was written for a man and the other for a woman. The spell for the man, who was called Pausanias, comes from a cemetery in Acanthus and belongs to the late fourth or early third

34 DTAud 264,265, 266,267, 268,269, 270, 271, 292?.

35 For a conspectus of the defixiones found so far in Macedonia, see Voutiras (n. 23), 1, n. 1.

Downloaded from https:/ Swansea University, on 22 May 2017 at 19:47:47, subject to the Cambridge Core terms of use, available at https:/



century B.C. 36 On one side of the tablet Pausanias binds down a woman called Sime, so that she may be kept from the shrine of Athena and so that Aphrodite may not be propitious to her, until she is prepared to do what he wants and embrace him

(KaraSei, /xe'xP 1 <*v IJavaaviai




irorjajj oaa Flavoavias

jSouAerai. «ai /iijrt

JXTJTI A<f>po8iTTj

IXecos avrfi,





evaxfj £i(i.r] Side A.2-6). On the other side, Pausanias binds down someone called Ainis, who might be either a man or a woman, keeping him or her from religious rites and from possessing anything beneficial, until he or she propitiates him (jrplv av Tlavaavlav IXaar/Tai Atvis Side B.4-5). Pausanias almost certainly wishes Sime to bestow her sexual favours on him. In view of the uncertainty that surrounds the sex of Ainis it is rather less clear what Pausanias expects in that quarter. This is by some considerable margin the earliest defixio designed to bring a woman to a man. 37 Such indications as there are would suggest that Sime was sexually experienced and sexually available. Pausanias is, accordingly, neither a sexually obsessed youth trying to seduce one of the local maidens, nor a man looking for a wife. The other defixio comes from Pella and is to be dated to between 380 and 350 B.C. 38 Its very early date and its location in a place that had hitherto seemed remote from the main cultural currents of the Greek world gives it an unusual significance. The agent is a woman asking that the marriage between another woman, Thetima, and a man, Dionysophon, should not take place, that he should marry no other woman, widow, or maiden, and that he should take no other woman to grow old with him than herself. The spell ends by calling for a miserable death for Thetima and a blessed and prosperous fate for the agent. What happened in Pella in the fourth century B.C. may have been quite different from what went on in the rest of the Greek world at that time, but the existence of the spell refutes the thesis that women did not in fact engage in erotic magic.

Athens and mainland Greece in the fourth and third centuries


There are a considerable number of defixiones from Athens from the fourth century B.C. directed at persons who run brothels or taverns or who work in these establish- ments as prostitutes. It seems likely that the purpose of the persons commissioning them was to eliminate competition. The example of these spells should make us cautious about assigning a wholly or even partially erotic motive to spells that seek to impose constraints on the sexual activities of a man and his partners. For example, an Attic spell from the fourth century that attempts to prevent a man called Aristokudes from achieving sexual union with any of the other women who are presented to him or any boy may not have been composed by a woman torn by sexual jealousy, but by one who did not wish her livelihood to be threatened or destroyed (Api[a]TOKv8ri KO.1

36 E. Trakosopoulou-Salakidou, ' KardSea/ioi

a-no r-qv AKav6o\

in A.-Ph. Christides and


R. Jordan (edd.), FXtliaaa xau fj.aycta: Kei/xeva amo rr\v ap\ai6ri]Ta (Athens, 1997),



4; D. R. Jordan, 'Three curse tablets', in D. R. Jordan, H. Montgomery, and E. Thomassen

(edd.), Magic in the Ancient World, Proceedings of the First International Samson Eitrem Seminar

(Norwegian Institute, Athens 4-7 May 1997) (Bergen, 1999), 115-24 provides an improved text.

It effectively confutes the thesis tentatively propounded by C. A. Faraone, The agonistic context of early Greek binding spells', in C.A. Faraone and Dirk Obbink (edd.), Magika Hiera:

Ancient Greek Magic and Religion (New York, 1991), 15, and repeated more confidently at CP90 (1995), 4, n. 14 that defixiones intended to win the sexual favours of someone 'represent some kind of hybrid flowering of a later, more complex magical tradition'. To sustain the thesis Simaetha's KaraSeois of Delphis would also have to be left out of consideration.



E. Voutiras, REG 109 (1996), 678-82; id. (n. 23), 1-7.

Downloaded from https:/ Swansea University, on 22 May 2017 at 19:47:47, subject to the Cambridge Core terms of use, available at https:/


r d j <f>a.vo(v)iAeva,s I avra)



I y^-nor'



OLXXTJV yvvai(Ka)

•nraiSa DTWti 78). 39 It is even clearer in the case of spells which seek to put an end to the relations enjoyed by a woman with more than one man that sexual jealousy is not the primary motive. We have in an Attic spell that attempts to sever the relationship enjoyed by a woman called Theodora with two named men and with any other men she has dealings with a likely instance of motivation that is primarily economic at work (DTAud68). 40 It is to be surmised that the person responsible for the spell was a courtesan jealous of her trade. Spells that seek to put an end to the sexual activity of a woman and mention no man create a strong presumption that the woman attacked made a living from selling her body. There is a slightly less strong presumption that the person responsible for the defixio is another woman whose livelihood is threatened. From Boeotia there is a spell, perhaps of the Hellenistic period but certainly not later, in which an attempt is made to put an end to the activities of a woman called Zoi's from Eretria. Since the intercourse {awovalrj) in which Zois participates, her pleasure (r/8ov(r))) and bottom (nvyiov) are singled out for mention and are consigned to Hermes and Ge along with her playing of the kithara, there is little room for doubt that she was a prostitute-cum-entertainer. 41 Another binding-spell from Boeotia that can hardly be later than the last-named spell tries to drive asunder a man and a woman and then apparently a further man from the same woman. 42 The only reason for thinking that the focus of the spell is on the woman is the mention in it of a second man. The name of the woman appears to be Antheira, which again suggests a courtesan. Since what the spell principally seeks to put a stop to is the love-making, kissing, intimate talk, and mutual affection


KTJ evvav

KTJ XdXr/aiv KTJ <f>i\r)oiv A6) of the man and Antheira, the

very strong likelihood is that Antheira is a courtesan. The identity of the person casting the spell is even more difficult to ascertain, but a woman is again the more likely candidate.

Two spells of the third century A.D. from the Athenian Agora

Of the twelve spells written by the same hand found in a well in the south-western corner of the Athenian Agora at least three are erotic. 43 One of them (no. 7) seeks to impose a chill on the sexual relationship between a man and a woman. The spell affords no clue to the identity of the parties involved. Numbers 8 and 9 are more in- formative in that respect: both spells are in essence directed at a woman called Juliana and seek in one case (no. 8) to put an end to a sexual relationship between Juliana and two men, and in the other (no. 9) to end a sexual relationship between Juliana and a man called Polynikos. Since Juliana is to be found in an fpyaarrjpiov (no. 8.5), she is a prostitute. The men whose affections for her the spells seek to cool are her clients or, from their point of view, her lovers. It is a reasonable presumption that the motives of

39 Faraone (n. 37), 14 speaks of its having been written by a jealous wife or fiancee. He translates it: '[I bind?] Aristocydes and the women who will be seen about with him. Let him not marry another matron or maiden.' For the necessary corrections, see Voutiras (n. 23), 57 with n.


40 For an improved text, see A lex sacra from Selinous, ed. M. H. Jameson, D. R. Jordan, and R. D. Kotansky, Greek. Roman and Byzantine Monographs 7 (Durham, NC, 1993), 130.

41 E. Ziebarth, 'Neue Verfluchungstafeln aus Attika, Boioten und Euboia', SB phil.-hist. Kl. 33(1934), 22.

42 Ibid., 1041-2, no. 23.

43 D. R. Jordan, Hesperia 54 (1985), 205-55.

Downloaded from https:/ Swansea University, on 22 May 2017 at 19:47:47, subject to the Cambridge Core terms of use, available at https:/



the person or less likely persons who cast spells nos. 8 and 9 were at least in part economic.


This section of the discussion may be concluded with several observations: (i) the authors of the formularies do not imagine that erotic magic is the exclusive preserve of men; (ii) the physical record does attest to women engaging in erotic magic; (iii) if we extend the notion of erotic magic to cover magic designed to further or impede making a living from selling sexual favours, then there is even more evidence of women being active in this sphere; (iv) the formularies and defixiones which survive from Late Roman Egypt, the defixiones from Hadrumetum in the High Empire and from Athens and Macedonia in the fourth and third centuries B.C. do nothing to encourage either the belief that those who practised erotic magic were youths unable to gain access to chaste and well-protected maidens or the thesis that erotic magic was practised by men trying to better themselves by a good marriage; (v) the evidence from Greece and Macedonia in the classical and early Hellenistic period suggests that erotic magic, broadly understood, was practised just as much by women as by men; the one man who can confidently be said to have practised erotic magic was interested neither in marriage nor in winning a carefully protected maiden for himself, but in getting a woman who sounds like a hetaera to submit to his desires.

Erotic magic in literature

The assertion that in Greek and Roman literature it is predominantly women who are portrayed practising love-magic requires careful examination. We need to see whether that statement is in fact true and then, if it is true, whether it means very much anyway. The latter task calls for an analysis of who exactly the persons portrayed performing love-magic are, what they are doing, and what the literary context of the portrayal is. Basically, what has to be determined is whether the men who wrote about women performing magic did so for ideological reasons or because they suffered from some form of false consciousness, however construed, or because they had a particular literary motive for doing so that varied from writer to writer. If literary motives are at work, then the observation that it is exclusively women who are portrayed performing erotic magic loses a good deal of its force. That is to say, if there are a variety of reasons for portraying women practising love-magic, then no single explanation should be advanced for all cases. I give a list of the women who are shown in imaginative literature and drama either practising love-magic or who are credited with an expertise in it followed by a list of male experts and users.

Women expert in erotic magic:

The Nurse in Eur. Hipp. The old women expert in incantations consulted by Simaetha at Theoc. 2.90-1. Niko the Thessalian sorceress in A. P. 5.205. Canidia and her companions in Hor. Epod. 5, 17 and in Sat. 1.8. Acanthis in Prop. 4.5. An unnamed saga in Tib. 1.2. Dipsas in Ov. Am. 1.8.

Downloaded from https:/ Swansea University, on 22 May 2017 at 19:47:47, subject to the Cambridge Core terms of use, available at https:/



The mother of Amyntas and Mycale in Nemes. Eel. 4. Possibly Philaenis in Mart. 9.29. Meroe at Apul. Met. 1.8-10, Pamphile at 2.5, 3.15-16, and the saga at 9.29-30. The Syrian sorceress at Lucian, Dial.meret. A A.

Women who make use of, or are prepared to make use of, erotic magic, though not them- selves experts:

Deianeira in Soph. Trach. Phaedra in Eur. Hipp. Simaetha in Theoc. 2. An unnamed woman in Verg. Eel. 8. Dido making a semblance of practising magic in Verg. Aen. 4. The wife of the baker at Apul. Met. 9.29-30. The courtesans Bacchis and Melitta in Lucian, Dial. Meret.

Men expert in erotic magic:

Jason at Pind. Pyth. 215-19. Moeris at Verg. Eel. 8. The Hyperborean magician at Lucian, Philops. 14. 44 Calasiris, who, because he is an Egyptian prophetes, is wrongly imagined to be expert in erotic magic (Heliod. Aeth. 3.16.2-3).

Men who make use of erotic magic, though not themselves adepts:

The lovelorn young man in Tib. 1.2. Mopsus and Lycidas in Nemes. Eel. 4. The rich young man at Lucian, Philops. 14. The young Thessalian hero who approaches Calasiris, on the false assumption that as an Egyptian priest he will be expert in erotic magic (Heliod. Aeth. 3.16.2-3).

There are indeed many more women than men on the list, though the ratio of roughly three to one is not quite as startling as one has been led to believe. The picture changes dramatically, however, if we compile a catalogue of the references to men and women engaging in love-magic in non-imaginative literature.

Women who are adepts in the use of erotic magic:

The stepmother of the plaintiff who persuades Philoneos' concubine that she has love-potions capable of restoring the affections of the latter (Antiph. 1.14,19). Lais, who gave love-philtres to her lovers to maintain her sway over them (Ar. Plut. 302-9 with E ad 303). Ninos, who, according to an ancient commentator, was accused by Menecles of making love-philtres for young men (Z in Dem. 19.281). The assertion in a declamation that the whole life of a prostitute is consumed in veneficium ([Quint.] Decl. maior. 14.5).

44 I have excluded from the list of men expert in erotic magic the Syrian stranger (Aoovpios £eivos) from whom Simaetha has learned the KCLKO. ^apjuaica with which she threatens to break her erstwhile lover Delphis, if her KardSeafioi do not bring him back to her (Theoc. 2.159-62). There is nothing to suggest that he is supposed, as Winkler ([n. 1], 228, n. 31) believed, to be expert in erotic magic and more knowledgeable than the old women from whom Simaetha had sought help.

Downloaded from https:/ Swansea University, on 22 May 2017 at 19:47:47, subject to the Cambridge Core terms of use, available at https:/



The old woman from whom a young man accused of giving a prostitute a love-philtre acquires that substance ([Quint.] Decl. Min. 385.6). The old women whom young men consult about love-magic (Philostr. V.A. 7.39). The old women found beside altars who teach rich women about love-philtres and incantations (Clem. Alex. Paed. 3.4.28,1.252 Stahlin).

This list could be expanded considerably, if we were to add to it the references in the Church Fathers and in John Chrysostom, in particular, to the use of erotic magic on the part of prostitutes.

Women who make use of erotic magic or avail themselves of the services of magicians for amatory purposes, but who are not themselves expert in the use of such magic:

The concubine who is persuaded by a wife that she has love-philtres in her possession capable of restoring a man's affections (Antiph. 1.14,19). A woman called Numantina who was tried in A.D. 22/23 for having used love-magic against her former husband (Tac. Ann. 4.22). Caesonia, the wife of Caligula, who by giving her husband an amatorium made him mad (Suet. Cal. 50). The advice that Plutarch proffers women in general against using ^iXrpa and other forms of yorjreCa to entrap men; he argues that men so ensnared have their senses damaged (Coniug. Prae. 139a). The rich Macedonian woman who took Alexander of Abunoteichos and his companion into her entourage, so that she might attract lovers (Lucian, Alex. 6). 45

Men expert in erotic magic:

The unspecified men in Rome who sell Thessalian philtres and incantations to women (Juv. 6.610-20). The doctor who had Alexander of Abunoteichos as his apprentice and Alexander himself (Lucian, Alex. 5-6). The educated Greeks who are hired men in rich Roman households (Lucian, De Merc. Cond. 40). The male magicians who help young men to perform love-magic (Philostr. V.A. 7.39). The mendicant-priests who along with old women are the source of instruction in love-magic for rich Alexandrian women (Clem. Alex. Paed. 3.4.29,1.252 Stahlin). The young man, apparently of good family, from Gaza who makes his way to Memphis to receive instruction in magic at the shrine of Asclepius there, so that he may return to Gaza to cast a spell on a girl who has resisted his approaches (Hieron. V Hilar. PL. 23.39). An Egyptian magician who teaches Appion an erotic incantation to help him win the woman for whom he longs {Clem. Horn. 5.4).

Men who resort to erotic magic, though not necessarily adepts:

The young men for whom Ninos was said to make love-philtres (S in Dem. 19.281).

45 She did so, although she was a bit over the hill, because she still wanted to draw men (i^wpov fiev. ipdopiov Se en etvai f}ov\ofievT]v). Lucian puts the same expression into the mouth of Melitta the hetaera who is trying to find an old women who knows how to perform incantations

that make women desirable to men: iira&ovoai KO.1 epaofiiovs


(Dial. Meret. 4.1).

Downloaded from https:/ Swansea University, on 22 May 2017 at 19:47:47, subject to the Cambridge Core terms of use, available at https:/



Callisthenes, the freedman of Lucullus, who gave Lucullus a love-potion, so that his love for him might be the more intense (Nep. fr. 1 Malcovati apud Plut. Luc. 43.1-2; cf. Plut. Plin. H.N. 25.25). The young men who are advised by Ovid that resorting to erotic magic is wrong (Ars Am. 2.99-106). The young man accused by a pimp of having given an amatorium to a meretrix that he had acquired from an old woman ([Quint.] Decl. Min. 385). The young men who both speak to old women about love-magic and get practical help from male magicians in engaging in it (Philostr. V.A. 7.39). Heresiarchs and their followers (Iren. Haer. 1.7.4,16.3). Appion, obsessed with desire for a woman, who receives both practical help from a Egyptian expert in magic and learns his spell {Clem. Horn. 5.3). Philometor, who used magic against the philosopher Sosipatra (Eunap. VS. 6.9). Dissolute Jewish youths from Tiberias who practised love-magic against respectable women (Epiph. Haer., P.G .41.416-20). The Syrian who conjured up a demon to take possession of a girl, only to be exposed

by the Holy Macedonius (Theodoret. Hist. Relig. 13.10-12)

A law-student in Beirut at the end of the fifth century A.D. who tries with the help of his associates to seduce a respectable Christian woman by magic (Zacariah, Vita Severi, P.O. 2.62-3).

The list could be expanded, but by this point it should be clear that counting heads is

a fruitless exercise. There is no evidence of an assumption on the part of non-

imaginative male writers that erotic magic was the special preserve of women and

that it was only practised by love-lorn females. Young men are believed to engage in it

in their pursuit of young women. They get help both from old women and male

magicians. No mention is made of older men resorting to such magic. Women are

also believed to have recourse to it, but they are either prostitutes or bored married women or perhaps widows. They, too, use either male or female experts. The one class

of women which is not generally credited with practising magic is that of respectable

unmarried girls. There is, accordingly, a striking asymmetry: prostitutes and married women practise magic on their own behalf; but only unmarried males engage in it. This may have been broadly speaking what happened, but the true picture can hardly have been so tidy. It looks, accordingly, as if the question that has to be asked is not why male writers imagine that only lovelorn women resort to erotic magic and have consequently chosen

only to depict such women practising magic, but why in imaginative writing it is mainly women who are depicted using love-magic. Before we address that question, it would

be prudent to look more closely at the instances on which the generalization has been

based as well as others by which it might be bolstered to see what kind of love-magic it

is in which women are portrayed engaging, what their motives are for resorting to it,

and whether they have scruples about using it or not. Deianeira in Sophocles' Trachiniae is hardly a love-lorn maiden resorting to erotic magic, but a married woman about to lose her husband to a younger woman. She uses what she imagines is a love-ointment only with the greatest reluctance and has strong moral scruples about what she does (582-7). Euripides' Phaedra, no maiden either, is certainly wracked by the agonies of unrequited love, but she does not at all like the suggestion of the Nurse that love-philtres and incantations should be used against Hippolytus and is extremely unhappy at being pushed into using them (Hipp. 486-9).

Downloaded from https:/ Swansea University, on 22 May 2017 at 19:47:47, subject to the Cambridge Core terms of use, available at https:/



Dido in Aeneid 4 is most definitely lovelorn. She has, accordingly, little difficulty in convincing her sister that the measures she is taking are designed to draw Aeneas back to her by magical means, but it is surely significant that to give verisimilitude to what she does she feigns reluctance at having to follow such a course of action (testor, cara, deos et te, germana, tuumque dulce caput, magicas invitam accingier artis 492-3). Simaetha, on the other hand, in Theocritus 2 displays no such scruples: when she first falls for Delphis, she approaches old women expert in performing incantations for help (91); then when he deserts her, she uses a binding-spell to try to bring him back (yvv Se viv €K Ovewv 10). The unnamed countrywoman in Vergil's eighth Eclogue, who has been deserted by the man she thinks of as her husband, is not credited with having any scruples about resorting to magic to bring the man back


In sum, the lovelorn woman who turns to magic to win the man she longs for certainly exists in literature, but there are only two straightforward examples of the type: Phaedra in the Hippolytus and Simaetha in Theocritus 2. Almost all of the other women in the catalogue use magic, because they have been deserted by their lovers, either to get them back or out of vengeance. They are: Deianeira in Trachiniae; Simaetha; the woman in Vergil, Eclogue 8; Dido in the Aeneid; Canidia in Horace, Epode 5; the saga in Apuleius who keeps an inn; and Melitta and Bacchis in Lucian. There is one predatory married woman: Pamphile in Apuleius. There are also old women in the Roman elegists who perform erotic magic on behalf of others. Overall, then, the lovelorn woman who uses magic to win the affections of a man turns out to be a not particularly commonfigure.The lovelorn men who have recourse to sorcery to win the women for whom they long are no fewer in number—in fact, three: Tibullus in 1.2; the rustic lovers in Nemsesianus 4; and the young man in Lucian's Philopseudeis.

It is conspicuously the case that many of the women who are portrayed engaging in

love-magic are prostitutes, ex-prostitutes, or incipient prostitutes. To this category belong Simaetha, Canidia, the sagae in the Roman elegists, the saga in Apuleius' Metamorphoses who runs an inn, and Melitta and Bacchis in Lucian. None of these women have the slightest scruple about having recourse to sorcery. Respectable women, on the other hand, are very reluctant to have anything to do with magic-working:

Deianeira, Phaedra, and Dido all express distaste for the course of action they are forced to follow or feign. The conclusion to be drawn is clear: prostitutes were at various times and places believed to practise magic freely, whereas in Augustan Rome and in Athens in the fifth century respectable women were imagined to be extremely unwilling to engage in such practices.

I turn now to the question of the motives writers have for portraying women

practising erotic magic. Those who subscribe to the view that there was a systematic distortion on the part of male writers and poets of the role played by women in erotic magic are in some sense committed to the theory that authors using very different literary forms all had the same motive for writing about female love-magic. In tragedy,

for instance, the theme of magic going wrong and leading to unintended and disastrous results lends itself to a form of drama based on the assumption that human beings very often take decisions that may have catastrophic consequences for themselves. Nor is any allowance made for the possibility that female magic-working may be a natural part of the subject-matter peculiar to the literary form. If in fact prostitutes and ex-prostitutes form part of the cadre from which in many ancient communities magic-workers were drawn, it is not surprising that they should also be presented in literary portrayals as actively engaged in magic and in erotic magic in

Downloaded from https:/ Swansea University, on 22 May 2017 at 19:47:47, subject to the Cambridge Core terms of use, available at https:/



particular. 46 The Roman elegists, for example, write as if they were young men deeply involved with courtesans and their world. That they should have encounters with lenae and sagae who practise magic hardly calls for comment. Nor is it surprising that a writer in the Greek east in the second half of the second century A.D. who invents conversations between prostitutes, partly inspired by Greek New Comedy, should make erotic magic one of the topics that engages the attention of his characters. This leaves us with the women in Apuleius' Metamorphoses who practise erotic magic on their own behalf: Meroe, who keeps a tavern; and Pamphile, the wife of Lucius' host in Hypata in Thessaly. They should not be treated separately from the other women in the Metamorphoses whose spectacular and horrific magic-working provide a good deal of the excitement in that work. The fact that Lucius the narrator has experience of female magic-working and hears from others about further instances of the same activity has a simple explanation: the tale is set in Thessaly, a land notor- ious for its sorceresses. Meroe and Pamphile are, in other words, one manifestation of a phenomenon wide- spread in Thessaly, female magic-working. They come, however, in forms recognizable to the Graeco-Roman world: the old prostitute who uses her knowledge of magic to take revenge on lovers who have rejected her or spoken ill of her, and the rich woman who practises magic in her pursuit of lovers. Both in Italy and in the Greek-speaking

east, taverns were well known as centres of prostitution.

taverns and used them as a base for procuring will themselves have been prostitutes; Meroe belongs to the type. Meroe and Panphile are, in sum, stock-figures but based on a reality; their magic-working has been greatly exaggerated by Apuleius or his source to make it suitably frightening for a Thessalian setting. In sum, the male poets, playwrights, and novelists who write about women practising erotic magic do so for a variety of reasons. The reasons that can be discerned sufficiently explain why the writers portray women practising erotic magic. Not only do writers have no one reason for depicting women performing erotic magic; the woman portrayed exhibit very different attitudes to such magic: some have no scruples about having recourse to it; others feel some anguish at being forced by circumstances to employ it. There is, furthermore, no one literary form in which erotic magic is portrayed. It is, accordingly, difficult to believe that one single factor is responsible for the choice of female erotic magic-working as a topic for literary portrayal.


Many of the women who ran


In a recent study of erotic magic Faraone maintains that men cast spells to arouse ipws in women, while women cast spells to retain or restore affection and esteem. 48 The latter category of spells are said to be called <f>iXrpa or ^api^aia, since they seek to secure <j>i\ia and to increase allure; the former are known as dywyal or <f>iXrpoKaTdheatJiOL. The two categories of spell in Faraone's view were thought to be quite distinct in their nature: those arousing epws were seen 'as a specialized exten-

46 On prostitution in antiquity the fundamental studies are H. Herter, RAC 3 (1957), s.v. Dime cols. 1154-1213, and id., JAC 3 (1960), 70-111, esp. 104 for magic and prostitution; on prostitution in the early Byzantine period there is Leontsini (n. 24). On prostitution and magic, see now Voutiras (n. 23), 56, n. 131.

47 Cf. T. Kleberg, Hotels, restaurants, cabarets, dans I'antiquite romaine, Bibliotheca Ekmania 61 (Uppsala, 1957), 89-91; V Vanoyeke, La prostitution en Grece et a Rome (Paris, 1990), 105-10. 4g C. A. Faraone, Ancient Greek Love Magic (Cambridge, MA, 1999).

Downloaded from https:/ Swansea University, on 22 May 2017 at 19:47:47, subject to the Cambridge Core terms of use, available at https:/


sion of cursing rituals and philia spells as a subcategory of healing and protective rites'. 49



Did these distinctions exist? Erotic defixiones and th e spells in the

labelled dycuyai or <f>i\TpoKaTa.8eofj,oi characteristically seek to induce both epws and

<f>iXia. 50 An d <j>CXrpa also seek

t o induce epws;

cf. th e Nurse's <f>lXrpa fioi


epa>Tos at Eur. Hipp. 509-10. 51 It is natural then for a formulary to speak about (friXrpa ipwTiKa (PGMXll.301-6). Not only women but also men use <f>(Xrpa or in their Latin form amatoria, and not just to secure the affections of an existing partner, but to win new lovers. 52 Pliny the Elder, for example, mentions amongst the substances supposed to function as amatoria" two specifically said to be effective in attracting women. 54 Finally, a wedge cannot be driven between spells designed to increase the x&P l s of a person an d those whose aim it is to arouse epcus. One of the feats that Lucian's magicians profess to be able to perform is to instil x°-P LS m matters of epcos (vviaxvovfievwv Kai xdpiras em TOIS epwriKots Alex. 5, De Merc Cond. 40). In the formularies the faXTpoKardSeofios or dyaty-q and the foXrpov have a good deal of overlap in the forms of expressions they employ. The most striking instance of the pattern is to be seen in a <f>iXrpov eTraivferov at PGM LXI.15-19 that seeks to impose exactly the same burning sensations, the same vertiginous feelings, and the same inability to eat and drink on its victim for exactly the same end as do dywyal and (jyiXTpoKardheafAoi. That is hardly surprising, since a (jyiXTpoKaTaSea/xos is only a specialized form of a faXrpov. 55 If a compiler of a formulary in his desire to impart a pseudo-precision to the spells he describes were to give a name to the procedures followed by Simaetha in trying to draw an errant lover to herself, he would surely have called them a ifiXTpoKardSeafios. That is in effect the way in which Simaetha herself describes what she does: vvv pev TOIS <f>iXrpois KaraS-qaofi-ai (Theoc. 2.159). The conclusion to which all of this points is that if there is any generic name for an erotic spell, it is <f>iXrpov and that dywyai and ^iXTpoKardSeafioi, to the extent that they are to be distinguished, are species of the genus <f>iXrpov. 56

University of Illinois at Chicago


49 Ibid., p. 30.




QiXTpoKardSeonoi and dyaiyai: PGM IV.351-2, 396, 1502, 1533-5, 2910, 2931-33, XVI.3-8, XXXVI.81,147,151-52; defixiones: XIXb.53-4, XXXVI.81.

Eur. Hel. 1102-4, Polyaenus, Strat. 8.38.1, Joseph A.J. 17.61-62, Ach. Tat. 4.15.4, 5.25.3 with 22.2, Porph. Abst. 2.41-2, Ioh. Chrys. Horn in I. Cor. 7:2 PG 51.216.

Men using ^iXrpa: E in Dem. 19.281, IG X.2.1.1026 = GVI1093; LSAM 20.15-22; Iren. Haer. 1.7.4, 16.3; Ach. Tat. 4.15.4; Porph. Abst. 2.41-42; Ov. Ars Am. 2.105-6; [Quint.] Decl. Min. 385.6. Winning new lovers: Ach. Tat. 5.25.3; Clem. Alex. Paed. 3.28.4. In the formularies under the rubric ^iXrpov the masculine gender is employed in prescribing how the spell is to be employed, which at the very least means that the use of such spells by men was envisaged (PGM VII.450,463, 661, XIH.237-9).


H.N. 25.25, 160,27.57, 125,28.34, 101,106, 30.141.


H.N. 28.101, 106.

55 So T. Hopfner, 'Philtron', RE 29 (1941), 208. The term faXTpoKardSeopos has the same form as <j>iXTponoaip.ov (Cyran. 1.18,22, 2.2,33, 3.7,37,38 Kaimakis), which is to say a <j>iXrpov -rroTifwv (PGM XIII.319). A <f>iXTpoirooip.ov may be used by a man to arouse Zpcos in a girl (Cyran. 1.18 Kaimakis); tfriXrpoTroTiofia employing the testicles o f a fox are effective against women when the left testicle is used and against men when the right is used (Cyran. 2.2 Kaimakis); much the same is true with the dove: the testicles work against women; the womb against men (Cyran. 3.37 Kaimakis).

56 In writing this paper I have incurred certain debts: to J. G. Howie and Alexander MacGregor for forcing me to clarify my thinking and for criticizing many infelicitous sentences; to David Jordan no t only for performing these services, but for his generosity in putting his great knowledge of Greek and Roman magic at my disposal.

Downloaded from https:/ Swansea University, on 22 May 2017 at 19:47:47, subject to the Cambridge Core terms of use, available at https:/